The present invention provides methods and systems for multiplexing five channels, such as 10 Gb/s to 50 Gb/s, into a single data sequence using a 5:1 multiplexer using a ⅕th ratio duty cycle clock. The ⅕th ratio duty cycle clock is a clock with a period equal to the channel data rate, and a pulse width equal to the period of data rate five times higher. The ⅕th ratio duty clock is combined with a proper combination of delays and phase shifters to allow the use of AND gates and OR gates to combine the five channels in a proper sequence to create a serial five-times higher data sequence.
|
10. A method of performing 5:1 multiplexing with a ⅕th ratio duty cycle clock comprising the steps of:
receiving a reference clock;
creating a ⅕th ratio duty cycle clock from the reference clock;
phase shifting the ⅕th ratio duty cycle clock to create five clocks, wherein each of the five clocks corresponds to a successive time periods; and
combining five data streams with the five clocks to create a multiplexed output.
1. A 5:1 multiplexer with a ⅕th ratio duty cycle clock, comprising:
a reference clock;
a clock module configured to create a ⅕th ratio duty cycle clock from the reference clock;
five phase shifters configured to phase shift the ⅕th ratio duty cycle clock by successive time periods to create five clocks, wherein each of the five clocks corresponds to time periods in a multiplexed output stream;
five AND gates, wherein the inputs to the five AND gates each comprise one of five data streams and one of the five clocks; and
a plurality of OR gates configured to combine the outputs from the five AND gates into a single data sequence.
2. The 5:1 multiplexer of
3. The 5:1 multiplexer of
4. The 5:1 multiplexer of
5. The 5:1 multiplexer of
6. The 5:1 multiplexer of
7. The 5:1 multiplexer of
8. The 5:1 multiplexer of
9. The 5:1 multiplexer of
11. The method of
12. The method of
13. The method of
|
The present invention relates generally to high-speed signal multiplexers, and more particularly to a 5:1 multiplexer with a ⅕th ratio duty cycle clock.
High-speed data rates common in networking environments, such as 100 gigabits-per-second (Gb/s), require multiplexers to create a serial stream from several lower-rate signals. At a 50 GHz operating frequency, for example, 2:1 and 4:1 multiplexers are known in the art. Such devices are configured with two or four inputs and a single, multiplexed output. Currently, 5:1 multiplexers are not available for high-speed applications, such as 50 GHz.
100 Gigabit Ethernet (100 GbE) is an Ethernet standard presently under early development by the Institute of Electrical and Electronics Engineers (IEEE). In late November of 2006, an IEEE study group agreed to target 100 GbE as the next version of the technology. Additionally, high-speed optical systems are moving beyond the existing 10 Gb/s line rates to higher time-division-multiplexed (TDM) rates, such as 40 Gb/s and 100 Gb/s. Advantageously, 100 Gb/s optical signals offer an effective way to achieve 1 Tb/s utilizing 10 wavelengths of 100 Gb/s signals over wavelength division multiplexing.
Disadvantageously, the lack of 5:1 multiplexers will require multiple cascaded multiplexers with several clocks to achieve a native 100 GbE or 100 Gb/s signal. Alternatively, 100 GbE could be transmitted in parallel, such as five 20 Gb/s signals. However, this solution is inefficient with respect to the transmission rate, requiring five transmitters and receivers per signal. A 100 Gb/s signal for optical transmission would not be transmitted in parallel because this would remove any advantage of having a higher TDM rate.
A 5:1 multiplexer could be formed using 4:1 and 2:1 multiplexers; however such an implementation requires three bit clock control which requires gating of three clocks. Disadvantageously, this gating is a very challenging approach. Additionally, existing 4:1 multiplexers at high data rates (e.g., 50 Gb/s and the like) utilize clock division internally, and therefore cannot be used in combination with a 2:1 multiplexer to create a 5:1 multiplexer.
No known native 5:1 multiplexers exist for high data rates, such as 50 Gb/s or 100 Gb/s. Thus, there exists a need for a 5:1 multiplexer operable at high data rates with a single reference clock.
In various exemplary embodiments of the present invention, five channels, such as 10 Gb/s to 50 Gb/s, are multiplexed to one with a 5:1 multiplexer using a ⅕th ratio duty cycle clock. The ⅕th ratio duty cycle clock is a clock with a period equal to the channel data rate, and a pulse width equal to the period of a data rate five times higher. The ⅕th ratio duty cycle clock is combined with a proper combination of delays and phase shifters to allow the use of AND and OR gates to combine the five channels in a proper sequence to create a serial five-times higher data sequence.
Advantageously, the present invention can implement a 5:1 multiplexer at high data rates with existing AND and OR gates. Also, the present invention allows multiplexing of data sequences to form, for example, a 100 Gb/s data sequence using two of the 5:1 multiplexers. This is used to form 100 Gigabit Ethernet or the like in a single data sequence, allowing optimized transmission as a single signal versus requiring multiple signals, such as 20 Gb/s. Further, the present invention utilizes a single reference clock to form the 5:1 multiplexed signal. The present invention can be integrated into a single application specific integrated circuit (ASIC).
In an exemplary embodiment of the present invention, a 5:1 multiplexer with a ⅕th ratio duty cycle clock includes a reference clock; a clock module configured to create a ⅕th ratio duty cycle clock from the reference clock; five phase shifters configured to phase shift the ⅕th ratio duty cycle clock by successive time periods to create five clocks, wherein each of the five clocks corresponds to time periods in a multiplexed output stream; five AND gates, wherein the inputs to the five AND gates each include one of five data streams and one of the five clocks; and a plurality of OR gates configured to combine the outputs from the five AND gates into a single data sequence. Optionally, the 5:1 multiplexer includes a D Flip Flop which is input with the single data sequence and clocked with a clock equal to five times the reference clock. The output of the D Flip Flop includes a non-return-to-zero (NRZ) data stream including the five data streams in a single multiplexed output. Alternatively, the 5:1 multiplexer includes an AND gate which is input with the single data sequence and a phase-shifted clock equal to five times the reference clock. The output of the AND gate includes a return-to-zero (RZ) data stream including the five data streams in a single multiplexed output. The clock module include a first clock source (F1) and a second clock source (F1*5), and wherein the first and second clock source are synchronized to the reference clock. The ⅕th ratio duty cycle clock is created by inputting the first clock source and a phase-shifted compliment of the first clock source into an AND gate. The inputs to the 5:1 multiplexer include signals at bit rates between 10 gigabits-per-second and 50 gigabits-per-second. Optionally, the 5:1 multiplexer is contained within an application specific integrated circuit.
In another exemplary embodiment of the present invention, a method of performing 5:1 multiplexing with a ⅕th ratio duty cycle clock includes receiving a reference clock; creating a ⅕th ratio duty cycle clock from the reference clock; phase shifting the ⅕th ratio duty cycle clock to create five clocks, wherein each of the five clocks corresponds to a successive time periods; and combining five data streams with the five clocks to create a multiplexed output. Optionally, the method further includes inputting the multiplexed output into a D Flip Flop with a clock signal equal to five times the reference clock, wherein the output of the D Flip Flop includes a non-return-to-zero multiplexed data stream. Alternatively, the method further includes inputting the multiplexed output into an AND gate with a clock signal equal to five times the reference clock, wherein the output of the AND gate includes a return-to-zero multiplexed data stream. Optionally, the method is performed by an application specific integrated circuit.
In yet another exemplary embodiment of the present invention, a multiplexing system includes a reference clock; ten data streams; two 5:1 multiplexers each input with the reference clock and five of the ten data streams; and a 2:1 multiplexer input with the output of each of the 5:1 multiplexers, and the reference clock. The 5:1 multiplexer includes a clock module configured to create a ⅕th ratio duty cycle clock from the reference clock; five phase shifters configured to phase shift the ⅕th ratio duty cycle clock by successive time periods to create five clocks, wherein each of the five clocks corresponds to time periods in a multiplexed output stream; five AND gates, wherein the inputs to the five AND gates include one of five data streams and one of the five clocks; and a plurality of OR gates configured to combine the outputs from the five AND gates into a single data sequence. In one application, the multiplexing system is utilized to form 100 gigabit Ethernet. In another application, the multiplexing system is utilized to form a 1 terabit-per-second signal.
The present invention is illustrated and described herein with reference to the various drawings, in which like reference numbers denote like method steps and/or system components, respectively, and in which:
In various embodiments of the present invention, five channels, such as 10 Gb/s to 50 Gb/s, are multiplexed to one with a 5:1 multiplexer using a ⅕th ratio duty cycle clock. The ⅕th ratio duty cycle clock is a clock with a period equal to the channel data rate, and a pulse width equal to the period of data rate five times higher. The ⅕th ratio duty cycle clock is combined with a proper combination of delays and phase shifters to allow the use of AND and OR gates to combine the five channels in a proper sequence to create a serial five-times higher data sequence.
Advantageously, the present invention can implement a 5:1 multiplexer at high data rates with existing AND and OR gates. Also, the present invention will allow multiplexing of data sequences to form, for example, a 100 Gb/s data sequence using two of the 5:1 multiplexers. This could be used to form 100 Gigabit Ethernet or the like in a single data sequence, allowing optimized transmission as a single signal versus requiring multiple signals, such as 20 Gb/s. Further, the present invention utilizes a single reference clock to form the 5:1 multiplexed signal. The present invention can be integrated into a single application specific integrated circuit (ASIC).
Referring to
The CLK4 is provided to a clock distribution module 17 which is configured to distribute CLK4 to the 5:1 multiplexer 10. Alternatively, the multiplexer 10 can be configured to receive CLK4 and perform the functionality of clock distribution module 17 internally to the multiplexer 10. CLK4 is input into five delay/phase shifters 18, 19, 20, 21, and 22 to create five clocks CLK5, CLK6, CLK7, CLK8, and CLK9, which correspond to successive time periods for the multiplexed data, D1 through D5. The five clocks CLK5, CLK6, CLK7, CLK8, and CLK9 are each delayed by one bit of output data D5-5 relative to the previous one with the delay/phase shifters 18, 19, 20, 21, and 22 configured to provide the appropriate delay.
The five clocks CLK5, CLK6, CLK7, CLK8, and CLK9 are each input into AND gates 23, 24, 25, 26, and 27 respectively. The data D1, D2, D3, D4, and D5 are also input into AND gates 23, 24, 25, 26, and 27 respectively with the clocks. After the AND gates 23, 24, 25, 26, and 27, the data D1, D2, D3, D4, and D5 is in the appropriate time slot for multiplexing by five. The remaining devices perform a combination function utilizing OR gates 28, 29, 30, and 31.
The outputs of AND gates 23 and 24 are combined in OR gate 28 to provide a signal which combines D1 and D2. The outputs of AND gates 25 and 26 are combined in OR gate 29 to provide a signal which combines D3 and D4. The outputs of OR gates 28 and 29 are combined in OR gate 30 to provide a signal which combines D1, D2, D3, and D4. Finally, the output of OR gate 30 and AND gate 27 are combined in OR gate 31 to provide a signal which combines D1, D2, D3, D4, and D5. The present combination of AND gates 23, 24, 25, 26, and 27 and OR gates 28, 29, 30, and 31 can be arranged in different combinations as would be readily apparent to those of ordinary skill in the art.
In one exemplary embodiment, the output of OR gate 31 is input into D flip flop 32 which is clocked with CLK. CLK is the output of clock F1*5 14, which is five times higher than CLK1. CLK is input to the D flip flop 32 through a phase shifter 34. The output, D5-5, of the D flip flop 32 is a signal including D1, D2, D3, D4, and D5 multiplexed five to one. The D flip flop 32 performs retiming to form a non-return to zero (NRZ) output D5-5. In another embodiment, the output of OR gate 31 is input into an AND gate 33 along with CLK through a phase shifter as another input. Advantageously, this embodiment performs retiming to form a return-to-zero (RZ) output D5-5.
As will be readily apparent to those of ordinary skill in the art, the various components can include AND gates, OR gates, phase shifters, flip flops, clock oscillators, and the like. The present invention can be integrated into an application specific integrated circuit (ASIC), field programmable gate array (FPGA), or the like. It is envisioned that the present invention will operate at high frequencies, such as 50 GHz and the like.
Referring to
CLK5 is created by phase shifting CLK4 with phase shifter 18. CLK5 and the first data stream, D1, are input into AND gate 23 to form D1*CLK5 which represents D1, but in the appropriate time slot for the 5:1 multiplexing. Accordingly, CLK6, CLK7, CLK8, and CLK9 are created by phase shifting CLK4 with phase shifters 19, 20, 21, and 22. Then these clocks, CLK6, CLK7, CLK8, and CLK9, and the respective data streams D2, D3, D4, and D5 are input into AND gates 24, 25, 26, and 27, respectively, to form D2*CLK6, D3*CLK7, D4*CLK8, and D5*CLK9, which represent D2, D3, D4, and D5, but in the appropriate time slots for the 5:1 multiplexing.
D1*CLK5 and D2*CLK6 are input into OR gate 28 to form D1-D2, which represents a combination of data streams D1 and D2. D3*CLK7 and D4*CLK8 are input into OR gate 29 to form D3-D4, which represents a combination of data streams D3 and D4. D1-D2 and D3-D4 are input into OR gate 30 to form D1-D2-D3-D4. Finally, D1-D2-D3-D4 is input with D5*CLK9 into OR gate 31 to form D1-D2-D3-D4-D5, which represents all five data streams in a single sequence. Finally, the output D5_5 is retimed up with either D Flip Flop 32 (i.e., for an NRZ signal) or AND gate 33 (i.e., for an RZ signal).
Referring to
Phase shifting is performed on the ⅕th ratio duty cycle clock to create five clocks corresponding each to respective slots in the multiplexed stream, as depicted in step 53. This step is done by feeding the same ⅕th ratio duty cycle clock into delays, phase shifters, or the like, each with the proper adjustment. The five data streams are combined with the five corresponding clocks to form the multiplexed signal, as depicted in step 54. This step can be done with a combination of AND gates, OR gates, and the like.
Referring to
Similarly, the 5:1 multiplexers 10 of the present invention could be utilized to combine data sequences to form other signal types, such as a 100 Gb/s signal or a 1 Tb/s signal for optical transmission. Advantageously, the present invention overcomes existing limitations where current multiplexers are defined only for multiples of 2, such as 2:1, 4:1, etc.
Although the present invention has been illustrated and described herein with reference to preferred embodiments and specific examples thereof, it will be readily apparent to those of ordinary skill in the art that other embodiments and examples may perform similar functions and/or achieve like results. All such equivalent embodiments and examples are within the spirit and scope of the present invention and are intended to be covered by the following claims.
Patent | Priority | Assignee | Title |
Patent | Priority | Assignee | Title |
7325024, | Dec 04 2003 | Intel Corporation | Adder circuit with sense-amplifier multiplexer front-end |
Executed on | Assignor | Assignee | Conveyance | Frame | Reel | Doc |
Jan 09 2007 | KERSHTEYN, BORIS | Ciena Corporation | ASSIGNMENT OF ASSIGNORS INTEREST SEE DOCUMENT FOR DETAILS | 018794 | /0392 | |
Jan 11 2007 | Ciena Corporation | (assignment on the face of the patent) | / | |||
Jul 15 2014 | Ciena Corporation | DEUTSCHE BANK AG NEW YORK BRANCH | SECURITY INTEREST | 033329 | /0417 | |
Jul 15 2014 | Ciena Corporation | BANK OF AMERICA, N A , AS ADMINISTRATIVE AGENT | PATENT SECURITY AGREEMENT | 033347 | /0260 | |
Oct 28 2019 | DEUTSCHE BANK AG NEW YORK BRANCH | Ciena Corporation | RELEASE BY SECURED PARTY SEE DOCUMENT FOR DETAILS | 050938 | /0389 | |
Oct 28 2019 | Ciena Corporation | BANK OF AMERICA, N A , AS COLLATERAL AGENT | PATENT SECURITY AGREEMENT | 050969 | /0001 | |
Oct 24 2023 | BANK OF AMERICA, N A | Ciena Corporation | RELEASE BY SECURED PARTY SEE DOCUMENT FOR DETAILS | 065630 | /0232 |
Date | Maintenance Fee Events |
Jun 30 2008 | ASPN: Payor Number Assigned. |
Feb 08 2012 | M1551: Payment of Maintenance Fee, 4th Year, Large Entity. |
Feb 24 2016 | M1552: Payment of Maintenance Fee, 8th Year, Large Entity. |
Apr 03 2017 | ASPN: Payor Number Assigned. |
Apr 03 2017 | RMPN: Payer Number De-assigned. |
Mar 02 2020 | M1553: Payment of Maintenance Fee, 12th Year, Large Entity. |
Date | Maintenance Schedule |
Sep 09 2011 | 4 years fee payment window open |
Mar 09 2012 | 6 months grace period start (w surcharge) |
Sep 09 2012 | patent expiry (for year 4) |
Sep 09 2014 | 2 years to revive unintentionally abandoned end. (for year 4) |
Sep 09 2015 | 8 years fee payment window open |
Mar 09 2016 | 6 months grace period start (w surcharge) |
Sep 09 2016 | patent expiry (for year 8) |
Sep 09 2018 | 2 years to revive unintentionally abandoned end. (for year 8) |
Sep 09 2019 | 12 years fee payment window open |
Mar 09 2020 | 6 months grace period start (w surcharge) |
Sep 09 2020 | patent expiry (for year 12) |
Sep 09 2022 | 2 years to revive unintentionally abandoned end. (for year 12) |